AI Over 40 Series - Week 11: How AI Literacy is Revolutionizing Solution Design

When everyone says it’s impossible, it’s time to rethink what AI can do

After our acquisition by Cherry Bekaert, I was tasked with finding quick wins to leverage our expanded service offerings. My first idea was to introduce a tax partner to a key client for a sales-and-use tax review. Simple on paper—but I quickly discovered the “hand-raise” approach most firms use means you only act when the client notices a problem.

I surveyed major tax-software vendors. Same answer: no proactive outreach until after an audit. We were caught in a classic Catch-22: clients won’t pay to discover issues they don’t know they have, and developing exploratory tools is too expensive for speculative use.

How an AI advisory board changed the game

Frustrated, I turned to my AI advisors. Although I’m a CPA by training, I’d never built a tax-nexus engine. My pitch to the AI team was simple: “Can we automate an economic-nexus analysis inside Dynamics Business Central?” We have the transactional data; we just lack the logic and rules. AI’s first gift was a rough design template. It asked me questions I hadn’t considered—like saving partial assessments, handling thresholds by jurisdiction, and flagging anomalies. These were features I would have only discovered during long testing cycles.

The power of “vibe coding”

I soon found myself “vibe coding”—describing desired behavior in plain English and having models generate Business Central AL code. I fed Copilot, ChatGPT, Claude, and Gemini my natural-language specifications:

“Write an AL function that pulls sales transactions for each state, compares them against economic-nexus thresholds, and returns a compliance summary.” Each model produced working code snippets. I then asked them to review one another’s output, suggest optimizations, and estimate development time for average and senior developers. All predicted roughly 60 hours of work—matching our in-house developer’s estimate.

The unexpected acceleration

Here’s where things got interesting: another developer on the team saw the design doc and code samples. Without instruction, they dove into “vibe coding” themselves and delivered a functioning prototype in under two hours. That prototype included interactive reports, saved assessments, and basic exception handling. This wasn’t production-ready code—it lacked error logging, security reviews, and UI polish—but it was enough to validate the concept with our client and gather feedback for iteration. What used to require a massive budget now looked like an affordable pilot.

What AI literacy really buys you

This experiment taught me four lessons:

1. Affordable exploration. Problems deemed too costly to prototype become manageable when you can speak AI’s language in natural terms.
2. Superior design. AI surfaces edge cases and best practices early, improving the completeness of your specifications.
3. Faster prototyping. Developers can riff on plain-English designs and AI-generated code, slashing initial build times from weeks to hours.
4. Peer contagion. Once one team member experiences “vibe coding,” they evangelize the approach, accelerating adoption across the organization.

AI literacy isn’t just about writing better prompts—it’s about reshaping how we analyze, design, and validate solutions.

Checking reality

To be clear: a two-hour prototype is not a full implementation. You still need:

• Production-grade error handling and security
• Comprehensive testing and documentation
• Integration with change-management processes

But rapid prototyping transforms the conversation. Instead of selling clients on a speculative project, you can show a working model, discuss real data, and co-design the final solution together.

Why this matters now

Many leaders assume AI agents will “just build it” if they ask nicely. The truth is that understanding AI’s capabilities and limitations—knowing how to craft clear natural-language designs and request iterative code reviews—is where the real leverage lies. Agents can automate routine tasks, but literacy unlocks creative problem-solving at scale.

If we skip literacy and rush to agency, we risk deploying fragile bots that don’t address the core business needs. When everyone around you insists something’s too complex to prototype, your AI-savvy team can prove them wrong.

Your Week 11 Challenge: Tackle your “unsolvable” problem

1. Pick a stubborn issue. Something colleagues say is too risky, costly, or complex to explore.
2. Define it in plain language. Write a concise problem statement as you would to a fellow team member.
3. Ask AI to design. Get edge-case checklists, data requirements, and architectural templates.
4. Generate and review code. Use Copilot, ChatGPT, Gemini, or Claude to produce sample code and suggest refinements.
5. Rapid-prototype. Have a developer or yourself assemble a basic proof of concept in hours.
6. Iterate from feedback. Use client or stakeholder input to refine the prototype.

Document your journey in a one-page “AI design playbook” to share with your team.

The bottom line

We’re entering a new era where “too expensive to prototype” is no longer a valid excuse. AI literacy changes the economics of innovation, turning speculative ideas into tangible pilots. The question isn’t if AI can help solve your toughest challenges—it’s which “impossible” problem you’ll prove solvable first.

This post is part of our “AI Over 40” series. It first appeared on LinkedIn: AI for the Over 40 – Week 11: How AI Literacy is Transforming How We Design and Develop Solutions

Next Up: Exploring how these rapid-prototype cycles evolve into fully managed solutions, and what organizational shifts make that transformation sustainable.

Read more AI and Copilot blogs.

Trending Posts